seasonal variation in the distribution and ...la friaje durante el período de aguas altas, fue un...

SEASONAL VARIATION IN THE DISTRIBUTIONAND ISOTOPIC COMPOSITION OF PHYTOPLANKTON

IN AN AMAZON FLOODPLAIN LAKE, BRAZIL

Variación estacional de la distribución y composición isotópicadel fitoplancton en un lago de inundación en la Amazonia, Brasil

PEDRO CARABALLO1, Ph. D.; BRUCE R. FORSBERG2, Ph. D.; ROSSEVAL G. LEITE2, Ph. D.1 Grupo de Biodiversidad Tropical. Universidad de Sucre, Sector Puerta Roja. Sincelejo, Colombia. [email protected] 2 Instituto Nacional de Pesquisas da Amazonia, Manaus, Brazil. [email protected]; [email protected] author: Pedro Caraballo, [email protected]

Received 8 August 2013, first decision 12 December 2013, accepted 18 March 2014.

Citation / Citar este artículo como: CARABALLO P, FORSBER BR, LEITE RG. Seasonal variation in the distribution and isotopic composition of phytoplanktonin an Amazon floodplain lake, Brazil. Acta biol. Colomb. 2014;19(2):291-304.

ABSTRACTTo evaluate the seasonal variation and isotopic composition of phytoplankton, water samples were collected monthly betweenOctober 2007 and November 2008 in Lake Catalão, a floodplain lake at the confluence between rivers Negro and Amazon.Analyses of total chlorophyll concentration and δ13C and δ15N isotopic abundances were made from particulate size fractionsof 30-60, 10-30 and <10 µm in the littoral, pelagic, and floating meadows regions. Chlorophyll concentration was found tobe inversely associated to lake depth, and high concentrations of chlorophyll in the floating meadows zone were significant.The fraction <10 µm was the most abundant representing in average more than 40 % of the particulate matter. The δ13C valueswere relatively constant during the study (-25.1 ‰ ~ -34.0 ‰), whereas the δ15N values showed strong variability (15.6 ‰ ~2.4 ‰), which has been attributed to the resuspension of sediments during mixing of the water column. Mixing associated tothe sudden drop in temperature during the rising water period was an important event in the trophic and isotopic dynamicsof the lake. Variations in chlorophyll content were generally associated with the dilution process, in which concentration wasinversely correlated to the water level, whereas abundance was directly correlated to the water level.

Keywords: Amazon, floodplain lake, size-fractionated phytoplankton, stable isotopes.

RESUMENCon el propósito de evaluar la variación estacional de la abundancia isotópica (δ13C e δ15N) del fitoplancton, muestreosmensuales fueron realizados entre octubre de 2007 y noviembre de 2008 en el lago Catalão, un lago de inundación en la zonade confluencia de los ríos Negro y Solimões, ubicado frente a la ciudad de Manaus (AM, Brasil). Análisis de la clorofila totaly evaluaciones de la abundancia natural de δ13C y δ15N fueron realizados en las fracciones partículadas de 30-60, 10-30 y <10µm en las zonas litoral, pelágica y de macrófitas acuáticas. La concentración de clorofila presentó una relación inversa conla profundidad del lago, siendo relevantes las altas concentraciones encontradas dentro del tapete de macrófitas acuáticas.La fracción <10µm fue la que presentó la mayor concentración de clorofila, representando más del 40% del material par-ticulado. Los valores de δ13C fueron relativamente constantes durante el período de estudio (-25,1 ‰ ~ - 34 ‰), mientras quela abundancia natural de δ15N presentó una amplia variación (15,6 ‰ ~ – 2,4 ‰), que fue atribuida al proceso de resuspensiónde los sedimentos en los procesos de mezcla de la columna de agua. En este contexto, el fenómeno de la mezcla asociada con

Artículo de investigaciónACTA BIOLÓGICA COLOMBIANA

Acta biol. Colomb., 19(2):291-304, mayo - agosto de 2014 - 291

la friaje durante el período de aguas altas, fue un evento im-portante en la dinámica trófica e isotópica del lago. En gene-ral las variaciones de la concentración de clorofila fueronasociadas al proceso de dilución, en el cual la concentra-ción es inversa y la abundancia es directamente relacionadacon el nivel del agua.

Palabras clave: Amazonas, lago de inundación, fitoplancton,isótopos estables.

INTRODUCTIONPhytoplankton is one of the main energy sources of auto-trophic energy for the ichthyofauna in various tropical fluvialecosystems (Araujo-Lima et al., 1986; Hamilton et al., 1992;Forsberg et al., 1993). It is the same with Central Amazonfloodplains where phytoplankton represents a great part ofthe carbon content of the main species of commercial fish,both in adult (Forsberg et al., 1993; Benedito-Cecilio et al.,2002) and larval stages (Leite et al., 2002) despite making uponly 2 % of the regional primary production (Melack andForsberg, 2001). Thus, phytoplankton is a key element in thefood chain of the fishing industry, and since fish is the mainsource of protein of the riverine population (Santos et al.,2006), the study of phytoplankton is critical to regionaldevelopment. Since phytoplankton is composed of several species withdifferent sizes and spatial-temporal distributions, densitymeasurements provide little information on the dynamics ofthese communities. Chlorophyll concentration, in contrast,is the integrated biomass estimation that best represents thedynamics and productivity of the phytoplankton communityin aquatic ecosystems (Wetzel, 2001).The Amazon floodplain lakes are very productive systems, aswas demonstrated by Schmidt (1973) in Lake Camaleão, andby Fisher and Parsley (1979) in Lake Calado. The phytoplank-ton productivity in these lakes is related to the availability ofphosphorus and nitrogen (Forsberg et al., 1988; Huszar et al.,2006), which in turn depends on variable fractions that comefrom the river and have different geochemical propertiesdepending on the local drainage basin. This chemical andhydrological variability creates difficulties in the study ofchlorophyll dynamics and phytoplankton production in thestudied environments. Among the few existing studies, it isworth mentioning the research conducted by Schmidt (1973)due to its duration and scope and, among other things, theinverse relation found between the water level of the lake andthe concentration of chlorophyll, a pattern frequently obser-ved in floodplain lakes (Carvalho et al., 2001). Research onchlorophyll concentrations in the Amazon has generallydemonstrated that the largest concentrations of phytoplank-ton usually occur in the dry season and the lowest con-centrations in the flood season, as shown in the followingstudies: Fisher (1978) on Lake Janauacá and Rio Solimões

(Amazon River), Lopes and Bicudo (2003) on lakes in theAcre river basin; Camargo and Myai (1988) on TrombetasRiver; Schmidt (1982) on Tapajós River; Ibañez (1998) onLake Camaleão, and Brito et al. (2014) on Lake Catalão. The concentration of chlorophyll in a lake varies accordingto the physiological state of the cells, the species compo-sition, the depth, and the environment studied (Santos et al.,1999). This might make the assessment of the availability ofphytoplankton for zooplankton based on chlorophyll notvalid, particularly in systems with significant limnologicalvariations over short periods of time. This assessment maybe further complicated because the phytoplankton includesnon-digestible forms and other forms that cannot beassimilated by the zooplankton (Runge and Ohman, 1982).Even so, chlorophyll has been found to be useful in the mo-nitoring of phytoplankton in aquatic ecosystems, as has beendemonstrated in lakes and rivers in North America (Carlsonand Simpson, 1996). Since phytoplanktonic populationshave different photosynthetic and demographic behavioraccording to their size, analysis of size-fractionated chlorophyllis a quick method (unlike cell count) to follow populationchanges more efficiently. The study of size-fractionated phytoplankton makes it po-ssible to investigate the contribution of each fraction to thecommunity biomass, in the population dynamics andparticularly in the phytoplankton-zooplankton relationshipin the water bodies (Rai, 1982; Romero and Arenas, 1990),and as a useful tool for understanding the herbivore trophicchain. The size classifications of planktonic organisms, inrelation to phytoplankton, have important physical and phy-siological implications. The smaller cells have a lower velocityof sedimentation and are more efficient in terms of nutrientassimilation, growth, breathing and photosynthetic rate(Greisberger et al., 2008). This clearly shows the relativeimportance of the different size fractions that integrate aphytoplanktonic community, particularly in an ecosystem suchas Lake Catalão, where a predominance of nanoplanktonicalgae is found, especially Cyanobacteria and Chlorophyceae,in the low and high water periods of 2006-2007, respecti-vely (Almeida and Melo, 2011). This predominance of theCyanophyceae in aquatic ecosystems is associated with arelative contribution of picophytoplankton (<2µm) in theaquatic primary production higher than 70 % (Richardsonand Jackson 2007). On the other hand, Romero and Arenas(1990) have determined that 57 % of primary productioncomes from the <10µm fraction. In turn, Ansotegui et al.(2003), who assessed a 60-80 % production of chlorophyllin the <8µm fraction suggest that the predominance of thisfraction is normal in oligotrophic systems, whereas a pre-dominance of microplankton occurs in eutrophic systems.The predominance of one or other fraction generally definesthe route through which the carbon flows to higher levels.When there is a predominance of the smaller fraction, themain route, according to Adame et al. (2008), is carbon

292 - Acta biol. Colomb., 19(2):291-304, mayo - agosto de 2014

Caraballo P, Forsber BR, Leite RG

recycling through the microbial loop in the photic zone. One of the most efficient ways of studying the assimilationof phytoplankton carbon by organisms from other trophiclevels is through the study of carbon (δ13C) and nitrogen(δ15N) stable isotopes. In fact, the assessment process wouldbe simple if there was not the methodological problem ofsorting the phytoplankton from the other seston particles.Thus, Hamilton and Lewis (1992) proposed the sortingof detrital material by means of centrifugation of the samplein a silica solution, which was not very successfully whenperformed by Calheiros (2003), but being later improved(Hamilton et al., 2005). However, so far this method has notbeen largely used. Most studies of the isotopic compositionof phytoplankton have analyzed all particulate organicmatter smaller than 50-60 µm, under the assumption thatthe isotopic value produced represents the phytoplanktonicpopulations in the environment. One alternative methodused in some studies is the measurement of isotopic ratio ofdissolved inorganic carbon (Marty and Planas, 2008).

The purpose of the present study was to investigate theseasonal and spatial variation of phytoplankton in LakeCatalão, including the variations in chlorophyll concentrationand isotopic composition (δ13C and δ15N) associated withdifferent fractions of phytoplankton particulate sizes.

MATERIAL AND METHODSStudy AreaLake Catalão is a floodplain lake in the final portion of landthat separates the rivers Negro and Amazon (Solimões),across from the city of Manaus, at 3° 10’ 04’’ S and 59° 54’45’’ W (Fig. 1). The lake is influenced by the Rio Negro at thestart of the flood period and by the Amazon at the end ofthe flooding and during the high water periods. During thelow water period the lake can be isolated from the two rivers,though it is usually linked to the Rio Negro by a small channel.The hydrologic balance of Lake Catalão is greatly influencedby the relative magnitudes of the inflows of rivers Solimõesand Negro, and can be described as a variable mixture of


Seasonal Variation in the Distribution and Isotopic Composition of Phytoplankton in an Amazon Floodplain Lake, Brazil

Figure 1. Map of the region including the Lake Catalão zone and the Manaus city. (modified from Neves dos Santos et al., 2008).

these chemically different sources (Brito et al., 2014; Almeidaand Melo, 2011). This variable mixture results in a very particu-lar temporal and spatial distribution of physical and chemicalcharacteristics of the lake. The limnological periods of low, flooding, flood and fallingwaters are well defined in the lake, as they are the direct resultof the dynamics of two large rivers with a predictableflooding pattern. Figure 2 shows the water level in the portof Manaus, which corresponds to the water level of the lake.On November 22, 2007, when the water level of Rio Negroin the port of Manaus was 18.9 meters, the water began toflow into the lake and when the water level was 25.6 metersin the Port, water from Rio Solimões began to flow into thenorthern part of the lake. The flood period occurred betweenJune and July, the falling water period between August andSeptember, and the dry period between October andNovember. In the dry period, there is no water flow into thelake from the adjacent rivers, and the lake area is reduced. In a study based on samples collected during the dry andflood periods, an inventory of 235 taxa of phytoplanktonicalgae was registered by Almeida and Melo (2011): thesealgae belonged to 10 classes, with the greatest diversity ofspecies found for class Chlorophyceae, and the greatestabundance Cyanophyceae, particularly the species Synechocystisaquatilis, Synechococcus elongatus and Planktothrix isothrix. Theseasonal dynamics of the phytoplanktonic biomass of CatalãoLake were characterized by Brito et al. (2014) as follows: a

period of low density (rising and high water period), with theaverage values of chlorophyll in the photic zone varyingbetween 3.4 µg/l to 13.5 µg/l, followed by the falling and lowwater phases when increase in the photic zone was observed,with average chlorophyll concentrations of 12.2 µg/l to 24.7µg/l. On the other hand, studies of reproductive ecology offish suggest that the areas of confluence between rivers ofclear or dark waters are used for the reproduction ofmigrating Characiform species (Leite et al., 2002). Therefore,the coast of Lake Catalão at Rio Solimões, and Lake Catalãoitself, is a region of special interest to the study of variousaspects of the trophic food webs.

Field SamplingSample collection was carried out every fifteen days, throughintegrated samples from the littoral, pelagic and floatingmeadows zones. The littoral region was defined as the areawith sparse vegetation and no macrophytes; pelagic region wasthe area of open waters; floating meadows region was the areain the middle of the floating macrophytes. In all cases 20 L ofwater was collected and taken to the laboratory in less thanthree hours. In the pelagic region a three inch wide PVC pipewith water-tight closing mechanism was used, which allowswater passage when the pipe is lowered down and is closedwhen the tube is pulled up. The depth to which the pipe wastaken down was determined in situ as three times Secchi depthvalue, thus ensuring sampling within the entire euphotic zone



Figure 2. Average electrical conductivity (bars) of the water column in the pelagic zone and depth of water in the Rio Negro (line), from November2006 until December 2007.

according to extinction coefficient values estimated by Brito etal. (2014). In the littoral region, water was sampled usingplastic buckets and care was taken to avoid collection ofsediments. For sampling inside the floating meadow, waterwas first filtered in the boat using a series of three filters asfollows: 500 – 300 – 120 µm, to remove macroinvertebrates,small fish and fragments of plant roots.In situ assessments of temperature, dissolved oxygen andelectric conductivity were performed in the three environ-ments studied with the use of a Thermo Scientific Orion5-Star Plus device calibrated before every measurement. TheOD polarographic sensor has a 0.1 – 0.01 mg/l resolutionand a relative accuracy of ±0.2 mg/L. All parameters in thelimnetic region were measured meter by meter to the bottomof the water column. In the other regions measurements weretaken at the surface and at the bottom.

Size-Fractionation of PhytoplanktonThree particle-size fractions were assessed for chlorophyll andcarbon and nitrogen stable isotopes based on the studies ofRai (1982) and Romero and Arenas (1990): the <10µm frac-tion, usually called ultraplankton nanoplankton, the fractionbetween 10-30 µm called nanoplankton, and the fractionbetween 30-60 µm and a 30µm mesh plankton. The < 2µmfraction (autotrophic picophytoplankton) was not sampledin this study because the major phytoplanktonic speciesfound in Lake Catalão were more than 2µm Cyanophyceae(Almeida and Melo, 2011). Twenty liters of water were filtered through a 60 µm mesh toremove the large zooplankton represented by Copepoda andCladocera. The values for the chlorophyll content obtainedin this fraction are called total chlorophyll content or totalpigment content. The second fraction was obtained in thelaboratory through the filtration of 10 liters using a 30µmmesh. Five liters were analyzed and the values are called 10-30 µm total chlorophyll content or <30 µm total pigmentcontent. The remaining five liters were filtered with 10 µmfilter paper and the values obtained are called <10 µm totalchlorophyll content or < 10 µm total pigment content.

Chlorophyll DeterminationThe concentration of total pigment content was determinedbimonthly in three different particle sizes (<10, 10-30 and30-60 µm) according to the extraction technique usingacetone at 90 % described in Golterman et al. (1978). Foreach sample 200 ml of filtered water with 60, 30 and 10 µmwere collected with the use of GF/F glass filters, which arethe most efficient devices for retaining algal cells. The filterswere kept in aluminum foil bags in the freezer at -20 ºC. Forthe extraction of chlorophyll-a, the filters were macerated inaqueous solution of 90 % acetone. The bulk was centrifugedat 3,500 rpm for 20 minutes and kept in a cool and darkenvironment for 24 hours. Then, absorption of the super-natant solution was determined in a quartz cuvette of 1 cm

optical path length, at 663 and 750 nm wavelengths, bymeans of a FEMTO 700S spectrophometer. Then, the so-lution was acidified with HCl 1N, and the readings wererepeated at the same wavelengths. Determination of the total pigment content or total chlo-rophyll content was provide, as described by Carlston andSimpson (1996), which comprises of all pigments anddegradation products that are absorbed at 665 nm. Theequation of Golterman et al., (1978) was used in this studyfor the calculations of total chlorophyll: B (µg/L)n = (106U.Ve) / (kc. Vf). The extinction coefficient used was 89.

Stable Isotopes AnalysisA total of 138 isotope analyses of carbon and nitrogen inphytoplankton were performed, including three size fractions(<60 µm, <30 µm and <10 µm) and three environments(littoral- lit, pelagic-pel and macrophyte-mac).Isotopic analyses were provided in 500 ml of water from theeuphotic zone previously filtered (previously burnt for 1h at450 ºC) with 60 µm, 30 µm and 10 µm mesh nets. Withreference to the difficulty reported by Hamilton et al. (2005)involving the presence of particulate matter different fromalgae in the samples, the criterion of Araujo-Lima et al.(1986) and Forsberg et al. (1993) was adopted, according towhich the samples filtered with a 60 µm mesh net forremoving zooplankton are representative of the amount ofphytoplankton that exists in the environment. The GFF filters with sample were dried at 60 ºC for 24 hours,placed in 5 ml Eppendorf pipette and sent for isotopicanalysis at UNESP’s Center of Stable Isotopes in Botucatu,São Paulo. The samples were double analyzed by massspectrometer (IRMS/EA), with an analytical error of 0.2 %.The 13C/12C and 15N/14N ratios were assessed using asreference the PDB standard (Pee Dee Belemnitella in SouthCarolina, USA) for carbon and the atmospheric nitrogen forthis isotope. The values are expressed in delta per thousand(δ‰), as the result of the equation:

δ13C or δ15N = Rsample = {(Rsample – Rstandard)-1} x 103

where R is the 13C/12C or 15N/14N isotope ratio for sampleand standard.One test of filtering efficiency was done three times, alwaysusing water from the limnetic region of the lake. The isotoperatios of phytoplankton lower than 60 µm, lower than 30 µm,lower than 10 µm, lower than 5 µm and lower than 0.2 µmwere analyzed. The data are presented as average values ±standard deviation (sd) and analyzed using Anova. The ho-moscedasticity was also tested.

RESULTSTemperature of Lake Catalão during the study was on average28.9 ºC ± 0.9, with maximum and minimum values of 32.3– 27.6 ºC. The hottest month was November, with an average



temperature of 30.03 ºC and the coldest month was May,with an average temperature of 27.9 ºC, with this value in-fluenced by a phenomenon called “friagem” (sudden drop intemperature) that affected all the region. The electricalconductivity variation in the lake during the four limnologicalperiods reflects the variation in the proportion of the mixtureof waters from Rios Negro and Solimões in the lake (Fig. 2).At the dry period the conductivity values are on average 193µS ±4.0. In December, the waters from the Rio Negro floodedinto the lake reduce the conductivity value to 34 µS ± 6.6 inApril, the period of major influence of Rio Negro. At the endof April the waters from Rio Solimões flowed in and theconductivity increased until reaching the typical values forthis river in May. In August these values increased when thefall of water level began, and the values obtained were againthose found in November 2007 at the dry season, though alittle lower because they were measured in the first week ofthe month. The results of assessments of total pigments in the threeenvironments studied are shown in figure 3. The averagevalue of chlorophyll content in the entire environment studiedin the lake was 10.42 µg/L, with the largest concentration(29.8 µg/L) detected in the pelagic region in December, atthe beginning of the flooding with waters from Rio Negro.The lowest concentration was assessed in the two months offlood, May and June, when, except for the floating meadowszone, the values were on average 2.8 µg/L. During the samemonths, the total chlorophyll content in the floating meadows

zone was 6.3 µg/L, which are the highest concentrations in thelake for this period. In the months of August and September,which correspond to the falling waters period, the concen-tration of chlorophyll increased, reaching the greatest valuesin November when the sampled period finished. It is im-portant to stress that there were no floating macrophytes inthe lake in the dry season, which explains why we do not haveresults for this environment during this period. The pelagicregion of the lake always showed the lowest pigment values,but the tendency of variation of pigment values was generallythe same in the three environments. Despite the higher value of chlorophyll content in December,at the beginning of the flooding of the lake, the highest con-centration occurred in the two dry periods, and the highestaverage value was found in 2007 (Fig. 3). The lowest con-centrations corresponded to the high waters period and theintermediate values were observed in the flooding and fallingwater periods. The fact that the highest concentrations ofchlorophyll were found in the floating meadows zone duringthe flooding and flood is significant. Therefore, the concen-trations detected in the littoral region were highly variableduring the whole study period, though they were the highestduring the 2007 dry season. On the other hand, this valuewas lower in the same period in 2008, corresponding to nearhalf the value found in the pelagic region.The three fractions studied in all the regions showed differentaverage values of chlorophyll content during the study period(p<0.05) as indicated in figure 4. The largest concentrations



Figure 3. Monthly total pigments concentration in samples filtered through a net of 60 µm at three environments in the Lake Catalão: Pelagic(squares), littoral (X) and floating meadows zone (triangle) vs. the depth of the Lake Catalão (line with �).

were found in the <10 µm fraction and the smallest con-centrations in the less than 30µm fraction. On the otherhand, no difference was detected between the concentrationsof pigments in the three studied regions.The seasonal variations in the percentage of the three chlo-rophyll fractions considered here (60-30, 30-10 and <10 µm)are shown in figure 5. It is important to stress the significantcontribution of the total chlorophyll content of the <10 µmfraction in the three environments, which were respectively51 %, 44 % and 41 % for the coastal region, pelagic regionand floating meadows zone. The highest values are asso-ciated with to the dry season during the months of October,November and December, the latter corresponding to thebeginning of the flooding. In this last period, the <10 µmfraction represented 87 % of phytoplankton in the coastalregion, 63 % in the pelagic region and 78 % in the floatingmeadows zone. The 60-30 µm fraction had an averagecontribution of 30 % and the average contribution of theintermediate fraction (30-10 µm) was 22 % of the totalchlorophyll content in the lake. This last fraction showed themost variable behavior throughout the year in the threeenvironments, with the values varying from 0.0 % in thecoastal and pelagic regions during the high and fallingwaters, to 63 % in the floating meadows zone s in the sameperiod. The 60-30 and 30-10 fractions were unstable in the

pelagic and coastal regions, and one of them even disa-ppeared for a few months.With respect to the floating meadows zone, the three sizefractions were present throughout the months of the studies,and the lowest participation of the <10 µm fraction wasfound in this environment.

Phytoplankton Stable Isotopes Analysis The average value of δ13C for all the fractions in the threeregions studied was -30.37 ‰ ± 5.79 with a maximum valueof -25.14 ‰ (in the floating meadows zone, in April) and aminimum value of -33.95‰ (pelagic region, in September).The general values of phytoplankton δ13C in the threeenvironments studied are shown in figure 6. There is a sig-nificant correlation (p = 0.02; r2 =0.62) between the δ13Cvalues of the pelagic and littoral regions (Fig. 6A), which doesnot occur between these variables and the values obtainedfor phytoplankton in the floating meadows zone. It is im-portant to stress that no macrophytes occurred during thedry season in the lake, and when they occurred at thebeginning of the flood, they represented a mere extension ofthe coastal and pelagic regions, with similar chlorophyllconcentrations. The maximum δ15N values for integrated samples weremeasured in the coastal and limnetic regions in the 2007 dry



Figure 4. Average concentration of total chlorophyll in the size fractions studied in the Lake Catalão: <10 µm, <30 µm and <60 µm.

season, and an average value of 14.88 ‰ was obtained. Highvalues were also observed in June (12.08 ‰) in the threeenvironments, and also in July (12.58 ‰) in the pelagic region,

after the first phenomenon of sudden drop in temperature thatoccurred in the lake that year. Excluding these three monthswhen the lake was formed by a mixture of waters from the ana-



Figure 5. Percentage contribution of size fractions of the total chlorophyll in the three environments studied. A: littoral; B: pelagic and C: floatingmeadows zone. Black (<10 µm); Grey (10-30 µm) and white (30-60 µm).

lysis, the average δ15N value for the phytoplankton was 8.32‰ ±1.58 sd. The δ15N data for the three studied environmentsare presented in figure 6B.On the other hand, the three size fractions studied showedno significant difference between them at the δ13C and δ15Nmeasured values. The δ15N values had a minimum value of-1,19 ‰ in the fraction <60 µm from the floating meadowszone in September, and maximum value of 6,37 ‰ in thefraction <60 µm from the littoral region, in August. In general,the δ15N values in the three environments and fractions studiedshowed great variability. The δ13C values for the three fraction showed a high value of-26,51 ‰ in the fraction <30 µm from the floating meadowsand a minimum value of -39,71 ‰ in the fraction <60 µm from

the floating meadows zone. Although without a significant di-fference in the global values, there is a difference higher than1 ‰ at δ13C between the values of floating meadows zone, whichare richer than the values of the coastal and pelagic regions.

DISCUSSIONAnalysis of data showed no difference between the total ofchlorophyll content in the three studied zones, excluding thedata concerning the macrophytes in the dry period when thishabitat did not exist. The abundance of pigments, associatedwith phytoplankton, in the aquatic macrophyte communitieswas surprising for a floodplain lake, although a similar situa-tion was observed by Zocatelli et al. (2012) in a lagoon of theBrazilian northeast. During rising, high and falling water



Figure. 6. A and B. Monthly values of �13C ‰ (A) and �15N (B) in the three studied environments of the Lake Catalão. Floating meadows zone(triangle); Littoral (X); Pelagic (squares).

periods, the concentrations of chlorophyll in the floatingmeadows zone were not different from those in the limneticand coastal regions. These regions are considered poor orwithout significance in terms of phytoplankton because theyare covered by floating vegetation. According to Melack andForsberg (2001) phytoplanktonic communities in lakes aremostly restricted to open waters where there is more light(according to Lewis et al., 2001). Due to the constant move-ment of water through the aquatic macrophytes, a great partof phytoplanktonic production remains in this region and isused to feed invertebrates. These authors affirmed that themacrophytes act like a filter that avoids the excessive growthof algae in the pelagic region. This explanation of phyto-plankton production in macrophytes is consistent with thefact that the concentrations assessed in the beginning of theflood period were similar to those found in the coastal andlimnetic regions (Fig. 3). The chlorophyll contents in the floating meadows zone mightchange the projection of the primary production of phyto-plankton made by Melack and Forsberg (2001), which isexclusively based on the production in open waters of thelakes. Brito et al. (2014) had already found that the pelagicregions of floodplain lakes are not always the most produc-tive in terms of primary production. This occurs because thesunlight in the lake is significantly reduced by the concen-tration of particles dissolved and in suspension, which reducethe photic zone to around 3.0 m. Thus, trees and macrophytescan be more able to capture sunlight than phytoplankton(Melack and Forsberg, 2001), unless phytoplankton growsamidst this vegetation cover, by taking advantage of inter-mittent illumination (Diehl, 2002) calls.Unlike the total chlorophyll contents, the analysis of thechlorophyll contents showed differences among the three sizefractions studied. Phytoplankton of <10 µm fraction waspredominant in the three zones during the whole studyperiod. This fraction, that includes the picophytoplanktonand part of the nanoplankton, is mostly composed of cyano-phyceae, chlorophyceae, bacteria and protozoa. Phytoplank-ton contents of <10µm fraction and had an average con-centration of 4.95±5.08 µg/L, which are almost double andtriple from the 30-60 and 10-30 fractions, respectively.However, during the flood period the ratio of the 30-60 frac-tion increased and prevailed even during the falling watersperiod. This behavior correlates to the values of nutrientsdetermined by Brito et al., (2014), where the highest concen-trations of nitrogen and phosphorus occurred during the dryseason, in contrasts with the behavior observed in temperatelakes where the <10µm fraction is larger in oligotrophicwaters and smaller in eutrophic waters (Adame et al., 2008). Therefore two important elements must be considered: 1. thedecreased concentration of nutrients during the flood perioddetermined by Brito et al. (2014) which would limit the growthof cells, particularly the larger cells, and is less efficient forthe assimilation of nutrients (Grisberger et al., 2008), 2. Accor-

ding to Lampert and Trubetskova (1986), the herbivorouszooplankton is more active in the less than 30 µm fraction,which was demonstrated by Caraballo et al., (2011) for twocladoceran species in an in situ lake experiment. As a result,we suggest that more than a limitation in nutrients, thephytoplanktonic communities in the lake are effectivelycontrolled by the zooplankton which is limited in the dryperiod due to the extreme turbidity of waters, which in turnreduces the filtration rate of Cladocera and constrains thepresence of species such as the Daphnia gessneri, as demons-trated in laboratory and in situ experiments (Caraballo andHardy, 1995).Carbon isotopic data of the three studied fractions (<10, <30and <60 µm), corroborate to the data of Araujo-Lima et al.(1986), Forsberg et al. (1993), Hamilton and Lewis (1992).No difference was detected between the isotopic values of thethree study fractions, although the average value of the ≤10fraction was higher in almost 2 ‰ than in the two other frac-tions. According to Burkhardt et al. (1999), evidence has beenfound on the difference between the size of algae and thetaxonomic groups, on global phytoplankton isotope fractiona-tion. Since it is unlikely that these groups of algae use differentsources of CO2, the isotope difference can only be explainedby different fractionation rates during photosynthesis.Except for the extreme values of δ13C found in the samplescollected in the floating meadows zone, the values tended tobe more negative in the dry period when there is a predo-minance of the <10 µm fraction and richer in the flood whenthe <10 µm phytoplankton rate is reduced. This tendencymay reflect changes in the δ13C of the CO2 fixed by thephytoplankton, associated to the seasonal variation in theisotope characteristics of the organic material metabolizedin the system. The lower value of δ13C for the phytoplanktonin the dry period may reflect the greater contribution of orga-nic material derived from phytoplankton for the communitymetabolism during this period. The volume of waters in thelake is reduced during the dry period, the influence of marginalhabitats is lower and the chlorophyll concentration reaches itspeak value. Since the δ13C of the phytoplankton is relativelynegative compared to other groups of aquatic plants (Forsberget al., 1993), this would also result in more negative values forCO2 and phytoplankton in the referred period. Amazon waters are usually supersaturated in CO2 (Richey etal., 2002), then, the effect of the limitation of carbon isotopefractionation is minimum. Carbon isotope variation observedin submerged plants such as algae depends mainly on theisotope characteristics of the CO2 fixed in situ by respirationrates and its δ13C depends on the isotope characteristics ofthe organic material metabolized by the aquatic biota(Mayorga et al., 2005). In general, a change in the specificcomposition of the phytoplanktonic community is not expec-ted to produce a significant alteration in the δ13C, which wouldbe the result of different assimilation of bicarbonate in face ofCO2 deficiency by the different species.



The highest values of δ15N were found in the samples fromthe first flooding period and in the subsequent flood, exactlyafter the “friagem” phenomenon that occurred in the lake inthe end of May (INPE, 2007). This behavior is associatedtotal mixing of the water column that occurred in these twoperiods, in November 2007, as a consequence of the windsand low water level, and in May as as result of a convectionprocess that resulted in the cooling of the water surface.During the second dry period studied, which was not asintense as the first, the δ15N values found were as low as theones found during the flood period. According to Calheiros(2003), a great variation in the phytoplankton δ15N mayoccur during the assimilation of the different forms ofnitrogen (N2, ammonium, nitrate and nitrite).The base line of the aquatic trophic network, defined as theisotope composition of carbon and nitrogen of their primarysources of food includes phytoplanktonic and non-phyto-planktonic sources that occasionally contribute to the secon-dary production (Perga and Gerdeaux 2006; Lehmann et al.,2004). This explains the great variability observed in the δ15Nof phytoplankton in Lake Catalão: the mechanisms that definethe isotope ratio of nitrate in the surface are a response to theprocesses of nitrate assimilation by the algae, denitrificationin the bottom and mixture of the water column (Lehmann etal., 2004). Consequently, during the periods of lake stratifi-cation, when the anoxic hypolimnion is defined, the δ15Nproduct of denitrification is enriched, and since the mixture ofthe water column re-suspends the substances of the anoxichypolimnion, enrichment of the δ15N of phytoplankton canoccur through the consumption of enriched ammonium(Perga and Gerdeaux, 2006). However, other processes canalter the δ15N of phytoplankton, including the direct N2 fixa-tion, mostly by cyanobacteria, microzooplankton grazing,protein hydrolysis and microbial decomposition. The latter istraditionally associated with the increment of 15N of residualorganic material, but this is far from being indisputable(Lehmann et al., 2004).

CONCLUSIONSThe relatively high values of total pigment encountered infloating meadows were surprising, considering the traditionalassociation of phytoplankton with pelagic environments only.This phytoplankton may be derived from periphyton commu-nities (diatomaceae and chlorophyceae), which are commonin this habitat and can become detached, forming an “acci-dental” plankton community. It may also represent pelagicphytoplankton which can be concentrated in macrophytebeds as lake water flows through them. Some authors have associated the inverse relationship obser-ved in most floodplain lakes between the water level and thephytoplankton concentration with seasonal variations inherbivore density and the availability of nutrients and light.Our more cautious analysis suggests that this relationshipmay reflect a simple process of dilution resulting from large

seasonal variations in lake volume associated with the riverineflood pulse. When variations in nutrient and phytoplanktonconcentrations are considered together with changes in lakevolume, the natural production in these systems may actuallybe greater at high water.

ACKNOWLEDGEMENTSWe thank the Instituto Nacional de Pesquisas da Amazôniaof Brazil (INPA) and the Universidad de Sucre of Colombiafor cooperation and logistic support. We also thank BarbaraRobertson and an anonymous referee for a critical review ofthe manuscript. Financial support was provided by FINEPand Petrobras S.A. through the PIATAM and CTPETRO –AMAZONIA Projects and a fellowship from CAPES/FAPEAM.

REFERENCESAdame MF, Alcocer J, Escobar E. Size-fractionated phyto-

plankton biomass and its implications for the dynamicsof an oligotrophic tropical lake. Freshw Biol. 2008;53(1):22-31. DOI: 10.1111/j.1365-2427.2007.01864.x

Almeida FF, Melo F. Estrutura da comunidade fitoplanctônicade um lago de inundação amazônico (Lago Catalão,Amazonas, Brasil). Neotrop Biol Conserv. 2011;6(2):112-123. DOI: 10.4013/nbc.2011.62.06.

Ansotegui A, Trigueros JM, Urrutxurtu I, Orive E. Size distri-bution of algal pigments and phytoplankton assemblagesin a coastal–estuarine environment: contribution of smalleukaryotic algae. J Plankton Res. 2003;25(4):341-355.DOI: 10.1093/plankt/25.4.341

Araujo-Lima CARM, Forsberg BR, Victoria R, Martinelli LA.Energy sources for detritivorous fishes in the Amazon.Science. 1986;234(4781):1256-1258. DOI: 10.1126/science. 234.4781.1256

Benedito-Cecilio E, Araujo-Lima C. Variation in the carbonisotope composition of Semaprochilodus insignis, a detriti-vorous fish associated with oligotrophic and eutrophicAmazonian rivers. J Fish Biol. 2002;60(6):1603-1607.

Brito J, Alves LF, Espirito Santo HMV. Seasonal and spatialvariations in limnological conditions of a floodplain lake(Lake Catalão), connected to both the Solimões andNegro Rivers, Central Amazonia. Acta Amaz. 2014;44(1):121-134.

Burkhardt S, Riebesell U, Zondervan I. Effects of growth rate,CO, concentration, and cell size on the stable carbonisotope fractionation in marine phytoplankton. Geochim.Cosmochim Acta. 1999; 63(22):3729-3741. DOI:http://dx.doi.org/10.1016/S0016-7037(99)00217-3

Calheiros DF. Influência do pulso de inundação na com-posição isotópica (δ13C e δ15N) das fontes primárias deenergia na planície de inundação do rio Paraguai(Pantanal – MS). Tese de Doutorado. USP - Centro deEnergia Nuclear na Agricultura; 2003. Accessed 13 Oct2009. Available URL: http://www.cpap.embrapa.br/teses/online/TSE05.pdf.



Camargo AF, Miyai. RK. Caracterização limnológica do lagoCuruça: lago de várzea do rio Trombetas (águas claras),Pará. Acta Limnol Brasil. 1988;2(1):153-180.

Caraballo P, Hardy E. Fluctuación diaria de las poblacionesde Daphnia gessneri HERBST y Ceriodaphnia cornuta SARS(Crustacea-Cladocera) en el lago Calado (Amazonas,Brasil). Boletín Científico INPA, 1995;3:79-96.

Caraballo P, Sanchez-Caraballo A, Forsberg B, Leite R.Crescimento populacional e análise isotópica deDiaphanosoma spinolosum e Ceriodaphnia cornuta (Crustacea:Cladocera), alimentadas com diferentes frações deseston natural. Acta Sci Biol Sci. 2011;33(1):11-19.

Carlson RE, Simpson J. A Coordinator’s Guide to VolunteerLake Monitoring Methods. North American LakeManagement Society; 1996.

Carvalho P, Bini LM, Thomaz SM, Oliveira LG, Robertson B,Tavechio WLG, et al. Comparative limnology of SouthAmerican floodplain lakes and lagoons. Acta Sci Biol Sci.2001;23(2):265-273.

Diehl S. Phytoplankton, light, and nutrients in a gradient ofmixing depths: Theory. Ecology. 2002;83(2):386–398.DOI: http://dx.doi.org/10.1890/0012-9658(2002)083[0386:PLANIA]2.0.CO;2

Fisher TR. Plâncton e produção primária em sistemasaquáticos da bacia da Amazônia Central. Acta Amazon.1978;8(4):43-54.

Fisher TR, Parsley PE. Amazon lakes: Water storage andnutrient stripping. Limnol Oceanogr. 1979;24(3):547-553. DOI:10.4319/lo.1979.24.3.0547

Forsberg BR, Devol A, Richey JE, Martinelli LA, Dos SantosH. Factors controlling nutrient concentrations in Amazonfloodplain lakes. Limnol Oceanogr. 1988;33(1):41-56.

Forsberg B, Araujo-Lima CARM, Martinelli LA, Victoria L,Bonassi JA. Autotrophic carbon sources for fish of thecentral Amazon. Ecology. 1993;74(3):643-652. DOI:http://dx.doi.org/10.2307/1940793

Golterman HL, Clymo RS, Ohnstad AM. Methods for physicaland chemical analysis of fresh waters. IBP Handbookn.º 8, 2.º ed. Blackwell Publications, Oxford; 1978.

Greisberger S, Martin E, Dokulil T, Teubner EK. A com-parison of phytoplankton size-fractions in Mondsee, analpine lake in Austria: distribution, pigment compositionand primary production rates. Aquat Ecol. 2008;42(3):379-389. DOI: 10.1007/s10452-007-9095-1

Hamilton SK, Lewis Jr WM. Stable carbon and nitrogenisotopes in algae and detritus from the Orinoco Riverfloodplain, Venezuela. Geochim Cosmochim Acta.1992;56(12):4237-4246.

Hamilton SK, Sippel SJ, Bunn SE. Separation of algae fromdetritus for stable isotope or ecological stoichiometrystudies using density fractionation in colloidal silica.Limnol Oceanogr: Methods. 2005;3:149-157.

Huszar VLM, Caraco NF, Roland F, Cole F. Nutrient–chlorophyll relationships in tropical– subtropical lakes:

do temperate models fit?. Biogeochemistry. 2006;79:239-250. DOI 10.1007/s10533-006-9007-9

Ibañez MSR. Phytoplankton composition and abundance ofa central Amazonian floodplain lake. Hydrobiologi.1998;362:79-83. DOI: 10.1023/A:1002926318409

INPE. Climanálise Boletim. On line versión. Accessed 22 May2007. Available URL: http://www6.cptec.inpe.br/revclima/boletim/

Lampert W, Trubetskova I. Juvenile growth rate as a measureof fitness in Daphnia. Func Ecol. 1996(5);10:631-635.

Lehmann MF, Bernasconi S, Mckenzie J, Barbieri A, SimonaM, Veronesi M. Seasonal variation of the δ13C and δ15Nof particulate and dissolved carbon and nitrogen in LakeLugano: constraints on biogeochemical cycling in a eu-trphic lake. Limnol Oceanogr. 2004;49(2):415-429.DOI: 10.4319/lo.2004.49.2.0415

Leite RG, Araujo-Lima CARM, Victoria RL, Martinelli LA.Stable isotope analysis of energy sources for larvae ofeight fish species from the Amazon floodplain. EcolFreshw Fish. 2002;11(1):56-63. DOI: 10.1034/j.1600-0633.2002.110106.x

Lewis Jr WM, Hamilton SK, Rodriguez M, Saunders JF III, LasiM. Foodweb analysis of the Orinoco floodplain basedon production estimates and stable isotope data. J N AmBenthol Soc. 2001;20(2):241-254.

Lopes MRM, Bicudo CEM. Desmidioflórula de um lago daplanície de inundação do rio Acre, estado do Amazonas,Brasil. Acta Amaz. 2003;33(2):167-212.

Marty J, Planas D. Comparison of methods to determinealgal δ13C in freshwater. Limnol Oceanogr: Methods.2008;6:51-63.

Mayorga E, Aufdenkampe AK, Masiello CA, Krusche AV,Hedges JI, Quay PD, et al. Young organic matter as asource of carbon dioxide outgassing from Amazonianrivers. Nature. 2005;436:538-41.

Melack JM, Forsberg BR. Biogeochemistry of Amazonfloodplain lakes and associated wetlands. In: McClainME, Victoria RL. Richey JE, editors. The biogeochemistryof the Amazon Basin. Oxford University Press, Oxford;2001:235-274.

Neves Dos Santos R., Ferreira EJ, Amadio S. Effect of sea-sonality and trophic group on energy acquisition inAmazonian fish. Ecol Freshw Fish. 2008;17(2):340-48.DOI: 0.1111/j.1600-0633.2007.00275.x

Perga ME, Gerdeaux D. Seasonal variability in the δ13C andδ15N values of the zooplankton taxa in two alpine lakes.Acta Oecol. 2006;30:69-77. DOI: http://dx.doi.org/10.1016/j.actao.2006.01.007

Rai H. Primary production of various size fractions of naturalcommunities in a North German lake. Arch Hidrobiol.1982;95(3/4):395-412.

Richardson T, Jackson G. Small Phytoplankton and CarbonExport from the Surface Ocean. Science. 2007;315(5813):838-840. DOI: 10.1126/science.1133471



Richey JE, Melack JM, Aufdenkampe AK, Ballester VM, LauraLH. Outgassing from Amazonian rivers and wetlands asa large tropical source of atmospheric CO2. Nature2002;416:618-620.

Romero MC, Arenas PM. Aporte relativo de diferentes frac-ciones fitoplanctónicas a la producción primaria y con-centración de clorofila, en la laguna de Chascomus(Prov. de Buenos Aires, Argentina). Rev Brasil Biol.1990;50(2):327-333.

Runge J, Ohman M. Size fractionation of phytoplankton asan estimate of food available to herbivore. LimnolOceanogr. 1982;27(3):570-576

Santos GM, Ferreira EJ, Zuanon JA. Peixes comerciais deManaus. Manaus: Ibama/AM, Provárzea; 2006.

Santos AC, Calijuri MC, Moraes EM, Adorno MA, Falco PB,Carvalho DP, et al. Comparison of three methods forChlorophyll determination: Spectrophotometry andFluorimetry in samples containing pigment mixtures andspectrophotometry in samples with separate pigments

through High Performance Liquid Chromatography.Acta Limnol Brasil. 1999;15(3):7-18.

Schmidt GW. Primary production of phytoplankton in thethree types of Amazonian waters. III. Primary productionof phytoplankton in a tropical flood plain lake of centralAmazonia, Lago Castanho, Amazonas, Brasil. Amazoniana.1973;4:379-404.

Schmidt GW. Primary production of phytoplankton in thethree types of Amazonian waters. V. Some investigationson the phytoplankton and its primary productivity in theclear water of the lower rio Tapajós (Pará, Braz).Amazoniana. 1982;7(3):335-348.

Wetzel RG. Limnology, lake and river ecosystems. ThirdEdition. Acad. Press; 2001.

Zocatelli R, Bernardesb M, Turcqc B, Campello R, AlbuquerqueAl, Sifeddineb A, et al., Composição da matéria orgânicalacustre como ferramenta na reconstrução paleoam-biental na Lagoa do Caçó, Maranhão, Brasil. GeochBras, 2012; 26(1):67-78.



seasonal variation in the distribution and ...la friaje durante el período de aguas altas, fue un...

Documents